Forecasting Lithium Mine Output Using Satellite Data

Abstract

This article examines whether remote sensed satellite data can predict lithium mine output. We use pixel classification of land cover as a measure for long-term mine expansion and nightly light emissions as a measure for short-term operational activities. We find a strong correlation between satellite data in one period and individual lithium mine production volumes in the following period. Our timely and cost-effective forecasting method of lithium mine production can help decisionmakers in companies that rely on lithium and policy makers who want to ensure adequate lithium supplies in their countries.

Keywords: Remote sensing, Mines, Lithium, Forecasting, Satellite data, Commodities